Zogby Redux?

The poll of polls

On October 21, 2013 I did this post about the curious results from the Rasmussen Reports polls since the founder of Rasmussen Reports, Scott Rasmussen left his post there to found a digital research group called the Rasmussen Media Group, LLC.  I had noticed, especially a couple of weeks before I did the post, how every single day, the Rasmussen poll of the president's approval rating, was an outlier [giving a high approval rating for the president] in Real Clear Politics compared to the other polls.  And I don't mean it was just an outlier by one or two points of the Real Clear Politics average of polls, but by between 5-7 points higher every day.

I jokingly wondered if Rasmussen Reports had been taken over by some left wing Democrats who were manipulating the methodology used when Scott Rasmussen had been there.   Now I am not joking but wondering if that is the case.  Because I did that post more than two weeks ago and the outlier position of Rasmussen has been there every day since I did that post.  So, for more than a month The Rasmussen Reports poll has been a high outlier of the president's approval rating.

Last month [October], while most polls had the president's approval rating dropping, Rasmussen actually showed that the president increased his approval rating by 1 point.

Could a coincidence last for more than a month in polling?   Today [Tuesday, Nov. 5, 2013], the Gallup poll for the first time has the president' approval rating under 40% at 39%.   Today in Rasmussen it is 48%. 

Let us look at the Real Clear Politics average of polls from their great website.


______________________________________________________________
President Obama job approval
Polling Data

Poll	Date	Sample	Approve	Disapprove	Spread
RCP Average	10/17 - 11/4	--	43.4	51.9	-8.5
Gallup	11/2 - 11/4	1500 A	39	53	-14
Rasmussen Reports	11/2 - 11/4	1500 LV	48	51	-3
GWU/Battleground	10/27 - 10/31	1000 LV	45	52	-7
Reuters/Ipsos	10/25 - 10/29	1660 A	40	52	-12
The Economist/YouGov	10/26 - 10/28	662 RV	44	54	-10
NBC News/Wall St. Jrnl	10/25 - 10/28	800 A	42	51	-9
FOX News	10/20 - 10/22	1020 RV	41	53	-12
CBS News	10/18 - 10/21	1007 A	46	49	-3
ABC News/Wash Post	10/17 - 10/20	RV	46	51	-5
CNN/Opinion Research	10/18 - 10/20	RV	43	53	-10

______________________________________________________________
As you can see from the results above while the Real Clear Politics average of the approval rating is at 43.4% with a negative 8.5% spread [disapprove vs approve]

Rasmussen on the other hand has a 48% approval rating [a 4.6% higher outlier] and a negative spread of only 3% [a 5.5% outlier from the RCP average].   If this happened only one or two days or even one or two weeks, you could write it off to being an anomaly.  But when this outlier position, always to the high side, has occurred every day for more than a month, it makes me question why. 

I am not a pollster or statistics expert, so I admit I could be all wet.   I just wish, like I opined in my previous post, someone like Nate Silver, the pollster and statistics guru of the New York Times [@fivethirtyeight on twitter], would do some investigation. 

At one time Zogby used to be a respected poll like Rasmussen, but with some curious results and a curious methodology used, it has become persona non grata in the world of polls.   The Real Clear Politics does not [and that has been for a long time] use the Zogby polls anymore.

Will Rasmussen possibly become Zogby redux?

______________________________________________________________
UPDATE:   I noticed today that the Rasmussen Reports poll has the president's approval rating down to 46%.   Finally down to that low [which is still above the low 40's that most of the other polls show but at least not a huge outlier like it has been for more than a month].   That is the lowest rating it has been in more than a month.  Maybe the Tales reporting on this has finally had an effect on the methodology used or more likely that is wishful thinking on The Tales part.  :-)     We will see if Rasmussen finally has gotten out of outlier status or if this is just a one day phenomena.